In the prior art, to better facilitate the user's use of an application App, many applications all support speech input. As such, the user only needs to input speech data to the App upon use, operations are simple and use is convenient.
Specifically, in the prior art, after the user inputs speech data to the App, the App needs to perform speech recognition for the speech data input by the user, then uses one or more patterns in each preset field to perform semantic parsing for a speech recognition result, and obtains semantic parsing information corresponding to the speech data in each preset field. That is to say, the semantic parsing information has the same semantics as the speech data so that the App directly analyzes the user's demands according to the sematic parsing information and thereby performs corresponding operations. For example, the user may use the speech data to request the application to invoke a certain application, or send a certain piece of verbal information, or query for a certain piece of information, or query and reserve a hotel room or book an airplane ticket, or the like.
However, in the prior art, when the semantic parsing information is parsed, at least one segment pattern of a certain preset field might be used. However, the segment pattern includes less information quantity and cannot accurately identify a corresponding field so that a reliability of the sematic parsing information belonging to the preset field is lower. In the prior art, since the reliability of the parsed sematic parsing information belonging to the preset field is not verified, the field to which the parsed semantic parsing information belongs is wrong, and the App performs operations irrelevant to the field of the speech data input by the user.